With the rapid development of biometrics, fingerprint recognition is increasingly used in business and people's daily life. Among them, how to enroll high-quality fingerprint images has become an important issue to improve the success rate of fingerprint recognition.
In the related art, fingerprint images are usually enrolled on the same finger for multiple times, and then image quality of each enrolled fingerprint image is calculated. Fingerprint images whose image quality exceeds a preset threshold will be enrolled into a fingerprint database for subsequent fingerprint recognition.
However, ridges of some people's fingers are not clear, and the image quality of fingerprint images acquired using the above method is relatively low and thus fails to reach the preset threshold, consequently, fingerprints of these people cannot be enrolled into the fingerprint database. The fingerprints of these people may be enrolled into the fingerprint database if the above mentioned preset threshold is decreased, which in turn cause that a large number of low-quality fingerprint images are enrolled into the fingerprint database, resulting in higher false reject rate and false accept rate.